Characterization of a rare IL-10–competent B-cell subset in humans that parallels mouse regulatory B10 cells by Yohei Iwata, Takashi Matsushita, Mayuka Horikawa, David J. DiLillo, Koichi Yanaba, Guglielmo M. Venturi, Paul M. Szabolcs, Steven H. Bernstein, Cynthia M. Magro, Armistead D. Williams, Russell P. Hall, E. William St Clair, and Thomas F. Tedder Blood Volume 117(2):530-541 January 13, 2011 ©2011 by American Society of Hematology

Enumeration of human blood IL-10-competent B10 and B10pro cells. Enumeration of human blood IL-10-competent B10 and B10pro cells. (A) Visualizing IL-10+ B cells. Purified blood mononuclear cells were cultured with BFA, LPS plus PMA, ionomycin and BFA (PIB), or LPS plus PMA, ionomycin, and monensin (PIM) for 5 hours and stained for cell viability, cell surface CD19 expression, and cytoplasmic IL-10. Representative cytoplasmic IL-10 staining by viable, single CD19+ B cells is shown in the flow cytometry dot-plots, with percentages indicating cytoplasmic IL-10+ B-cell frequencies within the indicated gates. Blood mononuclear cells that were cultured with BFA alone before immunofluorescence staining served as negative controls, with background staining similar to that obtained using isotype-matched control mAbs. Bar graphs represent mean (± SEM) B10-cell frequencies from 3 persons. (B) Representative IL-10 production by B cells from a person with relatively high B10-cell frequencies. B10 cells were identified after in vitro stimulation for 5 hours as in panel A. Alternatively, IL-10+ B-cell frequencies were determined after in vitro B10pro cell maturation by stimulation with LPS, CD40L + LPS, CpG, or CD40L + CpG, with PIB added during the final 5 hours of 48-hour cultures. As negative controls for IL-10 staining, only BFA was added to some cultures during the final 5 hours. Percentages indicate the frequencies of cytoplasmic IL-10+ B cells within the indicated gates among total CD19+ B cells. (C) B10-cell frequencies in persons after with TLR agonist stimulation as in panels A and B. Dots represent results from single persons after 5-hour culture with BFA alone, PIB, or the indicated TLR agonist + PIB. Horizontal bars represent means. (D) CD40L induced optimal B10 + B10pro cell maturation during 48 hours in vitro cultures with either recombinant CD40L or CD40 mAb, plus LPS for 48 hours, with PIB added during the final 5 hours. Bar graphs represent means (± SEM) from 5 persons. Similar results were obtained in 2 independent experiments. (E) Representative B10 + B10pro-cell frequencies after in vitro maturation and stimulation. Blood mononuclear cells were cultured for 48 hours with media alone or media containing CD40L, along with the indicated TLR agonists, with PIB added during the last 5 hours of culture. Significant differences between cultures with or without CD40L are indicated: #P < .05; ##P < .01. (C-E) Significant differences between means of controls and individual stimuli are indicated: *P < .05; **P < .01. Yohei Iwata et al. Blood 2011;117:530-541 ©2011 by American Society of Hematology

Human B10 and B10pro cells in (A) cord blood and (B) spleen and tonsil. Human B10 and B10pro cells in (A) cord blood and (B) spleen and tonsil. B10 cells and B10 + B10pro cells were identified after in vitro stimulation for 5 hours and 48 hours, respectively, as in Figure 1. Representative results are shown along with graphs indicating IL-10+ B-cell frequencies within persons. Cells cultured with BFA alone served as negative controls for background IL-10 staining. Significant differences between means of BFA controls and individual stimuli are indicated: *P < .05; **P < .01. Yohei Iwata et al. Blood 2011;117:530-541 ©2011 by American Society of Hematology

Human blood B-cell stimulation induces IL-10 transcription and secretion in vitro. Human blood B-cell stimulation induces IL-10 transcription and secretion in vitro. (A) Time course of IL10 transcript induction. Purified CD19+ B cells were cultured with media alone or CD40L + CpG for the times indicated, with IL10 transcripts quantified by real-time reverse-transcribed polymerase chain reaction analysis. Bar graphs indicate mean relative IL10 transcript (± SEM) levels in 6 persons. (B) B cells secreting IL-10 express IL10 transcripts. Purified blood B cells were cultured with PMA and ionomycin for 4 hours before CD19 staining and secreted IL-10 capture (left panel). Cell surface IL-10+ and IL-10− B cells were isolated using the indicated gates and subsequently reassessed for IL-10 secretion (right panels) before relative IL10 transcript levels were quantified by real-time reverse-transcribed polymerase chain reaction analysis. Mean fold differences (± SEM) for IL10 transcript levels from 3 different persons are shown, with transcript levels normalized so that the relative mean IL-10− B-cell value is 1.0. (C) Cell surface signals that regulate cytoplasmic IL-10 expression. Blood B cells were cultured with CpG, CD40L, and antiIgM Ab (IgM) as indicated for 48 hours with PIB added during the final 5 hours of culture. Representative frequencies of IL-10-producing cells are shown, with bar graphs representing mean (± SEM) percentages in 5 persons. (D) TLR agonists that induce IL-10 secretion. Purified CD19+ B cells were cultured with media alone, CD40L, or with TLR agonists and CD40L as indicated for 48 or 72 hours. IL-10 secreted into the culture supernatant fluid was quantified by ELISA. Bar graphs represent mean IL-10 (± SEM) concentrations from more than or equal to 4 different persons. (A-D) Similar results were obtained in 2 independent experiments. Significant differences between means of cells cultured in media alone and stimulated cultures are indicated: *P < .05; **P < .01. Yohei Iwata et al. Blood 2011;117:530-541 ©2011 by American Society of Hematology

Phenotypes of human blood and tissue B10 cells ex vivo and B10 + B10pro cells after in vitro culture. Phenotypes of human blood and tissue B10 cells ex vivo and B10 + B10pro cells after in vitro culture. (A) Representative cell surface phenotype of blood B cells cultured with LPS + PIB for 5 hours. (B) Representative cell surface phenotype of blood B10 + B10pro cells after stimulation with CD40L + LPS for 48 hours with PIB added during the final 5 hours of culture. (C) Representative cell surface phenotype of spleen B10 cells cultured with CpG + PIB for 5 hours. (D) Representative cell surface phenotype of spleen B10 + B10pro cells after stimulation with CD40L + CpG for 48 hours with PIB added during the final 5 hours of culture. (A-D) Cultured cells were stained for viability and cell surface molecule expression, permeabilized, stained with anti–IL-10 mAb, and analyzed by flow cytometry. Representative cell surface molecule expression by IL-10+ (thick line) and IL10− (thin line) CD19+ B cells from 3 persons. Shaded histograms represent isotype-matched control mAb staining. Yohei Iwata et al. Blood 2011;117:530-541 ©2011 by American Society of Hematology

Ex vivo blood B10 and B10pro cells share cell surface markers with memory B cells. Ex vivo blood B10 and B10pro cells share cell surface markers with memory B cells. (A) Blood B10 cells predominantly exhibit a CD24hiCD27+CD48hiCD148hi phenotype. Purified blood B cells were cultured with CpG + PIB for 5 hours before immunofluorescence staining for viability, cell surface molecule expression, and cytoplasmic IL-10. Cell surface CD24, CD27, CD38, CD48, and CD148 expression by IL-10+ (thick line) and IL-10− (thin line) CD19+ cells was assessed by flow cytometry. (B) Cytoplasmic IL-10 induction does not affect the cell surface phenotype of B cells. CD19+ blood B cells were cultured with media on ice (thin line) or with CpG + PIB (thick line) for 5 hours before immunofluorescence staining and flow cytometry analysis as in panel A. (A-B) Shaded histograms represent isotype-matched control mAb staining. Results represent those obtained for 3 persons. (C) Distributions of B10 cells within B-cell subsets defined by CD24, CD27, IgD/CD38, and IgD/CD27 expression. Purified blood B cells were cultured with LPS + PIB for 5 hours before immunofluorescence staining and flow cytometry analysis as in panel A. The horizontal and vertical lines on each contour plot are shown for reference, with the lower left quadrants delineating the IgD−CD38− and IgD−CD27− subsets determined by control mAb staining. Results represent those obtained for 5 persons. (D) The ex vivo CD24hiCD27+ B-cell subset includes the majority of B10 cells. Purified B cells were cultured with LPS + PIB for 5 hours before immunofluorescence staining for cell surface CD19, CD24, and CD27 expression and cytoplasmic IL-10 expression, with subsequent flow cytometry analysis. (E) B10pro cells derive from the CD24hiCD27+ B-cell subset. Purified blood B cells were sorted into the CD24hiCD27+ and CD24lowCD27− B-cell subsets, as indicated by the gates shown with purities more than 90% when reanalyzed by flow cytometry. The purified B cells were cultured with CD40L plus either LPS or CpG for 48 hours, with PIB added during the final 5 hours of culture before the relative percentages of IL-10+ B cells within the indicated gates was determined. Similar results were obtained in 2 independent experiments. (F) Ex vivo CD24hiCD27+ B cells are the predominant source of secreted IL-10. Purified blood B cells were sorted into the CD24hiCD27+ and CD24lowCD27− B-cell subsets as in panel E and cultured with the indicated stimuli for 72 hours. IL-10 secreted into the culture supernatant fluid was quantified by ELISA. Bar graphs represent mean IL-10 (± SEM) concentrations from triplicate ELISA determinations. Significant differences between means from CD24hiCD27+ and CD24lowCD27− B cells are indicated: **P < .01. Differences between means from cells in media or with stimuli are indicated: ##P < .01. (G) B10-cell proliferation in vitro. Blood mononuclear cells were labeled with CFSE and cultured with CD40L and CPG (top panels) or CD40L and LPS (top panels) for 48 to 96 hours, with PIB added for the last 5 hours of culture. Histograms (right) represent CFSE expression by the IL-10+ (thick line) or IL-10− (thin line) B-cell subsets. Results are representative of 2 independent experiments. Yohei Iwata et al. Blood 2011;117:530-541 ©2011 by American Society of Hematology

B10-cell regulation of innate immunity. B10-cell regulation of innate immunity. (A) B10 cell effects on mitogen-stimulated T-cell cytokine production. Purified blood CD24hiCD27+ or CD24lowCD27− B cells were stimulated with CD40L plus CpG for 24 hours, isolated, and then cultured with CD3 mAb-stimulated CD4+ T cells for 72 hours. After PMA plus ionomycin stimulation, CD4+ T-cell TNF-α expression was assessed by flow cytometry (heavy lines). CD4+ T cells cultured alone are shown as positive controls (thin lines). Background cell staining using unstimulated T cells is shown (shaded lines). (B) B10 cells regulate monocyte cytokine production. Purified blood CD24hiCD27+ or CD24lowCD27− B cells were stimulated with CD40L plus CpG for 24 hours and were cultured with blood monocytes for 20 hours before cytoplasmic TNF-α expression by CD14+ monocytes was assessed after 4 hours of LPS stimulation (heavy lines). Anti–IL-10 mAb was added to some cultures as indicated (dashed lines). Monocytes cultured alone are shown as positive controls (thin lines), with background cell staining using unstimulated monocytes shown (shaded lines). (A-B) Results represent those obtained in more than or equal to 2 independent experiments. Yohei Iwata et al. Blood 2011;117:530-541 ©2011 by American Society of Hematology

Blood B10-cell frequencies in patients with autoimmune disease. Blood B10-cell frequencies in patients with autoimmune disease. (A) Representative B-cell cytoplasmic IL-10 expression by control (Ctrl) persons, and lupus (SLE), rheumatoid arthritis (RA), Sjögren syndrome (SjS), blistering skin disease (BD), and multiple sclerosis (MS) patients with relatively high B10-cell frequencies after in vitro CpG + PIB stimulation for 5 hours. B10 + B10pro cell maturation was induced by 48-hour CD40L + CpG stimulation, with PIB added during the final 5 hours of culture. Percentages indicate IL-10+ B-cell frequencies among CD19+ B cells. (B) IL-10+ B-cell frequencies as in panel A, with each dot representing single persons. Horizontal bars represent group means. The solid horizontal lines represent means plus 2 SD (95% confidence interval) for controls; and dashed lines, means plus 2 SD for all values. The patients are described in Table 1. Significant differences between means of patient groups and healthy controls are indicated: *P < .05; **P < .01. (C) Relative frequencies of B10 cells and B10 + B10pro cells identified for control persons and patients with autoimmune disease as in panel B compared after CpG or LPS stimulation, with each dot representing a person. (D) Relationship between cytoplasmic IL-10 expression levels and B10 + B10pro-cell frequencies in control persons and patients after stimulation with CD40L + CpG, with PIB added during the final 5 hours of 48-hour cultures. Linear mean fluorescence intensities (MFI) for IL-10+ and IL-10− B cells were determined using the gates indicated in panel A with the values shown representing a ratio of IL-10+ to IL-10− MFIs. A linear regression line (± 95% prediction bands, dashed lines) is shown for reference. Yohei Iwata et al. Blood 2011;117:530-541 ©2011 by American Society of Hematology